-We are using the single genome sequencing (SGS) technology we developed previously to analyze and understand the accumulation of genetic variation in gag/pol and env. We have made significant advances in additional assay development and have extended studies to a number of different patient groups, including chronically infected patients, both naive and on therapy, as well as in primary and early human immunodeficiency virus (HIV) infection (in collaboration with Margolick, Daar, and Kottilil), and in long-term nonprogressors (Mens). As a result, we are obtaining a more comprehensive picture of HIV genetic variation in vivo in the presence or absence of drug resistance.-We have expanded analytic approaches to HIV population genetics using SGS and we have developed new technologies. The SGS approach, as developed in the DRP, is rapidly becoming the standard approach to investigate HIV populations, with a number of groups and large networks employing the technique, notably the Center for HIV/AIDS Vaccine Immunology (CHAVI). We continue to investigate the utility of the approach, and expand applications. We have collaborated with M. Jordan to compare HIV population structure as determined by SGS or standard cloning methodologies. The results demonstrate concordance between methods but also identify certain discrepancies requiring additional study. We have also collaborated with W.-S. Hu in an in-depth investigation of intersubtype recombination, demonstrating adaptive effects at distant sites. As resistance to integrase inhibitors increases, and NIH clinics are enrolling more such patients, we are preparing to extend SGS to study the integrase sequence as well.-The DRP is also developing new technologies to investigate HIV-1 genetic variation. We are investigating massively parallel pyrosequencing techniques to study HIV population genetics. Although such ultra-deep technology has been used to study HIV-1, the utility of the approach remains uncertain, because it is not clear whether the approach can accommodate a highly genetically diverse virus population and yield accurate phylogenetic data. The DRP has an extensive database of single genome sequences from a large cohort of well characterized patients. These single genome sequences will provide the gold standard to compare results of pyrosequencing and determine the utility of massively parallel sequencing in genetic analysis of HIV-1 populations. We have also developed useful quality control procedures. In initial studies we identified improvements that are essential to prevent assay-induced recombination; these optimization procedures enable pyrosequencing to be used reliably to investigate recombination and epistasis in genetically diverse populations.-Understanding of the expansion of genetic diversity following infection from a genetically limited to a highly diverse population has useful implications for applicability in understanding the HIV epidemic. Based on our understanding of genetic variation in acute and chronically infected individuals, we developed a new bioinformatics algorithm to discriminate between recently and chronically infected individuals based exclusively on population-based commercial genotyping data. Development of this algorithm has yielded the invention report EIR #238-2009. Field testing is currently in development, and we anticipate that this technique will be of broad epidemiologic utility in investigating incidence rates of HIV-1 infection. We have obtained additional support for these studies (Bench to Bedside Award 2010 with Margolick and Daar). In collaboration with J. Brooks (Public Health Inst. Canada), we are expanding these studies to investigate acute infections in Canada, and in collaboration with M. Jordan we are also in the process of expanding these studies to investigate incidence rates in developing countries. In addition, bioinformatic approaches are also being developed to investigate correlations between the mutations identified by SGS and standard commercially available genotypes and phenotypes.-We have expanded the SGS approach to investigate cell-associated HIV RNA and DNA. We have developed a new technique to determine whether individual cells are infected with one or more than one provirus (supported in part by Bench to Bedside Award, 2004). Such single cell sequencing has revealed that the majority of infected cells are infected with a single provirus. Currently we are subjecting the data to rigorous statistical analysis to estimate the rate of dual and multiple infection. As multiply infected cells are the source of phenotypically mixed viruses that form the substrate for recombination, these data will yield useful information regarding the role of recombination in the spread of new mutations, including those conferring drug resistance. -The development of these techniques has led to new insights in HIV population dynamics in understanding the effects of antiretroviral therapy, the nature of replication in natural suppression of HIV, and population dynamics of non-subtype B HIV populations.-The nature of HIV-1 populations in patients undergoing antiretroviral therapy remains uncertain, and we are conducting an extensive genetic analysis of HIV-1 before and after initiation of antiretroviral therapy (Completed Protocol 97-I-0082, new Protocol 08-I-0221). These results will yield new information regarding the nature and timing of genetic bottlenecks occurring during antiretroviral therapy. Analysis of HIV-1 sequences at relatively low viremia has been limited by technical issues in amplifying the relatively few HIV-1 sequences present in plasma during therapy. We have successfully adapted the SGS procedure to obtain acceptable numbers of sequences from patients suppressed on antiretroviral therapy. In collaboration with M. Polis and D. Persaud (NIH Bench to Bedside Award, 2006) we are analyzing genetic variation in patients enrolled in Protocol 97-I-0082 (now 08-I-0221; F. Maldarelli, PI) who have been suppressed on antiretroviral therapy for prolonged (greater than 8 y) periods. Initial analyses demonstrate that HIV does not undergo a genetic bottleneck upon initiation of antiretroviral therapy; despite 100-10,000 fold decline in levels of peripheral viremia, no significant decreases in genetic diversity were detected in the first 1-2 y of therapy. These data indicate common source of virus infecting short lived cells (responsible for greater than 90-99% of virus produced prior to therapy) and longer lived cells (responsible for virus produced 1-2 years after therapy is initiated). After prolonged therapy, emergence of predominant clones (as previously noted by Bailey et al.) was detected in the majority (7/8) patients. These data suggest that the non-clonal populations slowly decayed over time or that the clonal population increased by cellular expansion. We are also applying population genetics approaches to quantify the emergence of drug resistance mutations in rebound viremia in patients undergoing antiretroviral therapy. We are specifically investigating the relative roles of mutation and selection in development of resistance to AZT and NNRTI, as well as quantifying the role of APOBEC mediated mutations in the emergence of the M184I mutation conferring resistance to 3TC, FTC, and ddI.[Corresponds to Project 3 in the October 2011 site visit report of the Clinical Retrovirology Section, HIV Drug Resistance Program]